A Potted History of Downhole Seismic Sources
One of the best things about writing your own blog is that you have complete content control, for better or for worse ! I have therefore decided to delay a chat about some of the many ways we can use a sonic log, and instead focus for the next couple of weeks on some downhole technology. Specifically, downhole seismic sources; looking at the history of their development and their use today primarily in crosswell seismic tomography but also in single well imaging and reverse VSPs.
As this is relatively recent history, a lot of the pioneers in this technology are still active and might even end up reading this so I apologize in advance for any misrepresentation and would welcome corrections or additions if so desired.
My main motivation for these articles is that of all the borehole seismic technology we have in our arsenal it is this which I think has been most underutilized over the years. As we will see next week there are many domains in which the use of downhole seismic sources can make a valuable contribution to the understanding of the subsurface; from oil and gas, to engineering, to mining and geothermal, and many different applications within those domains, from subsurface reflection imaging for structure, to velocity tomography for thermal, or EOR applications such a CO2 and steam front monitoring.
The promise of crosswell seismic is resolution. As both source and receiver are downhole, ray paths are short and do not pass through the highly attenuative surface layers and therefore the valuable high frequencies are retained. Crosswell seismic typically produces images of structure and reservoir properties between wellbores at a resolution ten times that of typical surface seismic data. Imagine your 30Hz surface seismic at 300Hz! Achievable? Yes, but there are a few constraints.
Obviously two wells are required for the duration of the project which could last several days and they need to be prepared appropriately. The wells need to be ‘close enough’ (more on that later) and in the same plane; vertical or horizontal, as this is a 2D (or 4D) survey.
Planning and modeling are required not only to confirm geophysical objectives but for budgetary purposes. Spatial resolution comes with seismic frequency, which requires non aliasing, which requires appropriate receiver and source spacing, all of which requires acquisition time and hence rig time (time two !).
Geometry of crosswell seismic acquisition setup
So, back to how close do the wells need to be? Guess what. It depends. It depends on the power of the source; how much time is available to stack data at each station and of course the properties of the material between the wells.
Over the years geophysicists and engineers have faced some very stringent challenges associated with downhole source design. Surface source design is easy (he said flippantly!), make it bigger and use more of them. Unfortunately, those are not approaches available downhole.
Characteristics required of a downhole source
Slim enough to fit inside wellbore
Robust enough to withstand downhole pressure and temperature regime
High energy output, repeatable with short cycle time
Operation and wellbore safety and efficiency
Commercial interest in downhole seismic sources ramped up in the late 1980’s. Hardee et al (1986) cites oil reservoir evaluation and the newly conceived ‘Continental Scientific Drilling program (CSDP)’ as drivers for development. Paulsson (1988) cites recognition from the DOE of the need for improved reservoir characterization techniques.
Chen et al (1990) was one of the earlier studies which describes a field test of a reverse VSP comparing a downhole airgun, water gun, explosive charge and perforating gun each recorded into a spread of surface receivers. Tests were conducted at two sites. Results from the first test site indicated that the water gun and airgun gave similar results as did the perforating gun and the explosive charge. For the second test the water gun and perforating gun were dropped in favor of the cheaper and more easily operable airgun and explosive charge.
Comparison of explosive and airgun downhole sources (Chen et al 1990). The airgun source was shown to generate significantly more tube waves
The explosive source resulted in clear direct P-wave arrival and at least two detectable upgoing reflections. The airgun source suffered from significant tube wave contamination, this is discussed in theoretical detail by Winbow (1991). Suffice to say that we now know that active coupling to the wellbore is crucial to provide a radiation pattern which will maximize the P and Sv components and minimize tube waves.
Perhaps however the most significant conclusion of this research was the final observations,
“Neither the air gun nor the explosive causes any significant damage to an open hole or new steel casing. However, both the explosive and the air gun may deteriorate the cement bond when operated in a cemented cased hole. The degree of damage depends upon the integrity of the cement and other borehole conditions. “
It is this concern about wellbore integrity that likely condemned further development and the commercial use of the majority of implosive sources downhole. Interestingly however one could argue that this very type of source has been the most utilized in borehole geophysics in the last 15 years or so. Perforation shots are now routinely utilized in microseismic monitoring of hydraulic fracturing in unconventional basins. Knowing the location of a perforation shot and being able to record it on downhole geophones (and sometimes surface arrays) is very useful for calibrating velocity models used in microseismic event location.
There is however one type of implosive source that continued to be used and has seen further refinement and development since its introduction in the 1980’s and that is the ‘Sparker’. This type of source employs two electrodes immersed in fluid which are charged through a capacitor which create a spark between them when discharged. The resulting vaporized fluid generates a bubble pulse which acts as the source which is characteristically broad band and highly repeatable.
There are historically several challenges with this type of source; the power output is less than other sources, it is fluid coupled resulting in significant tube wave generation, and the repeat time between shots tends to be long (whilst the capacitors recharge), the electrodes also typically need replacing relatively frequently.
Heigl et al (2012) discuss a downhole sparker optimized for single well imaging and reverse VSPs for which the frequency output can be adjusted such that the majority of the bubble energy was concentrated in the first several hundred Hertz. Results showed transmission distances of over 4600ft.
Mode of operation of Sparker source (Heigl et al 2012)
Common receiver gather of fifty shots with sparker set to 100 Hz and receiver being about 4650 ft away (Heigl et al 2012)
One way to mitigate the potential wellbore damage of an implosive source is to use a ‘swept’ source where the source energy is dissipated into the formation over a period of seconds and thus significantly reduces the threat of any wellbore damage. Such sources include Piezoelectric, hydraulic and electromagnetic vibrators.
Comparison of time domain power distribution of impulsive and controlled frequency sources (Winbow 1991)
R&D (and associated funding) of downhole seismic source projects continued throughout the 1990’s. There was a lot of research into using piezoelectricity as a basis for a downhole seismic tool. The reverse piezoelectric effect is the conversion of electrical energy into mechanical energy such that ceramic plates can be induced to vibrate with specific frequencies based on the applied electrical current. These tools are fluid coupled and typically utilized hydrophone receiver arrays.
Chapman et al (1991) use this setup to unambiguously show evidence for anisotropy at the BP Devine test site. As part of the Stanford Tomography Project (STP), Harris et al (1992) discuss a similar acquisition set up acquire crosswell seismic data to image a carbonate reservoir in West Texas.
Dodds et al (1994) describe the acquisition of six tomography profiles as part of a project for UK Nirex Ltd at Sellafield, Cumbria, UK. The site was a potential location for an underground repository for Nirex to store nuclear waste. The hardrock of the granite subsurface was an ideal medium for the propagation of energy from one wellbore to another. Borehole pairs were between 300ft and 2000ft apart, and data was acquired with source and receiver level spacing of between 15ft and 20ft. The piezoelectric source used generated a central frequency of 1000Hz and resulted in a vertical resolution of approximately 15ft. A velocity tomogram from one of the well pairs is shown. Whether as a consequence of sub surfaces images such as this, or not, we will never know, but in 1997 the Secretary of State for the Environment in the UK rejected Nirex's case.
Piezoelectric sources are still used today. The high frequency of the signal (several kHz) can be advantageous for closely spaced well pairs resulting in high resolution tomograms. However, the relatively low energy output is restrictive and the absence of lower frequency content (less than about 100Hz) can make data interpretation and comparison with surface seismic problematic.
P-wave velocity tomogram between BH 2 and BH 4 (Parker et al 1994)
Concurrent with the R&D of Piezoelectric based sources was development of both hydraulic and electromagnetic based sources.
Paulsson et al (1996) provides an update on an initiative started some years earlier with a project group which had then grown to include Conoco, Amoco, Chevron, Exxon and DOE (Sandia). The source was a three-component hydraulic seismic vibrator deployed on custom fiber optic wirelines. It was a high-power source (peak output 7000 lbs) designed specifically for large interwell distances of up to 6000ft. The source was clamped to the wellbore and utilized a hydraulically oscillating reaction mass either axially or radially to generate orthogonal component data rich in P, Sh and Sv data. The source was rated to 200degC and 12,000psi with a frequency bandwidth of between 10 and 1000Hz. The Sandia report publically released in 1998 (Cutler 1998) interestingly recounts early discussions with service companies in 1993. Apparently, they were uninterested in a technology which had no commercial market, in the case of Schlumberger it took them another 16 years to be convinced as in 2009 they bought Z-Seis. Another interesting comment from the Sandia report was the service company’s reluctance to take on any technology that did not run on standard seven conductor cable. Some things never change!
Seismic source and receivers in cross well configuration (Paulsson et al 1996)
Many years later, and in contrast to the hydraulic vibrator described above, Nalonnil et al (2013) provide an update on an electromagnetic swept source specifically designed for longer offsets. The Z-trac source is magnetically clamped to the wellbore and can generate programmable sweeps between 30 Hz and 600 Hz. It is worth noting the lower ‘low’ frequency end when compared to the Piezoelectric sources, which are useful when comparing data with surface seismic. The sweep is highly repeatable with a small time interval between sweeps, data can therefore be stacked to improve the signal to noise ratio of the data. The radiation pattern of the source is directional, typically two modules are deployed together offset by 90 degrees to record direct S-wave and P-wave data. Usefully, the source can be run on a standard seven conductor cable.
Dual wavefield data set with P-wave data left and S-wave data (right). (Nalonnil et al 2013)
Lastly and by no means least was the work that Conoco were undertaking during the 1980’s. They had their own inhouse development program, which was working on, amongst other things, a distinctly different type of vibrator source. The orbital vibrator source (OVS) was to generate shear waves from a fluid-filled borehole without the direct mechanical coupling of borehole seismic sources to the borehole wall. It consisted of an encased spinning eccentric mass which when activated creates centrifugal forces resulting in compression on one side of the mass and tension on the other. This compression is in part converted to Sv and Sh polarized energy at the borehole wall with the same primary frequency as that of the spinning mass, resulting frequencies are between 70 Hz and 400 Hz.
The main challenge with this source is understanding the source characteristics (Nakagawa 2004). I also wonder about potential casing damage with the source spinning eccentrically, ‘bouncing’ off the sides off the wellbore, a little like when a washing machine on spin becomes unbalanced.
Borehole orbital vibrator source (left) and cross-sectional view (right)
Today the commercial availability of downhole seismic sources has consolidated. Vendors such as Avalon Sciences Ltd in the U.K. offer a sparker, the AST-1. Geodevice also offer several models of Sparker. The z-Trac source is still commercially available from Schlumberger as are their X-series piezoelectric sources. From what I know, and having had a look online is seems that other main vendors and service companies do not offer a downhole source. As always though I am happy to be proven wrong!
Of course, as for most things in the borehole seismic domain the future if fiber, specifically DAS, and it looks like the same might be true for downhole sources. Paulsson et al (2019) discusses an innovative high temperature, high pressure, single well system for geothermal applications. I am certainly looking forward to seeing a case study on that.
Next week I will be looking at how the use of downhole sources can make a valuable contribution to the understanding of the subsurface; from oil and gas, to engineering, to mining and geothermal, and many different applications, from subsurface reflection imaging for structure to velocity tomography for thermal or EOR applications such a CO2 and steam front monitoring.
Be positive and test negative !
References:
S. T. Chen, E. A. Eriksen, and M. A. Miller, (1990), "Experimental studies on downhole seismic sources", GEOPHYSICS 55: 1645-1651.
Robert P Cutler, 1998, “Development of a hydraulic borehole seismic source” Sandia Report, SAND98-0932, April 1998
Dodds, Kevin & Chouzenoux, Christian & Conn, P. & Emsley, S.. (1994). “Planning, equipment and execution of crosswell surveys.” 10.3997/2214-4609.201409881.
G. J. Elbring, H. C. Hardee, and B. N. P. Paulsson, (1989), "A test of a controlled downhole seismic source", GEOPHYSICS 54: 1193-1198.
H. C. Hardee, G. J. Elbring, and B. N. P. Paulsson, (1987), "Downhole seismic source", GEOPHYSICS 52: 729-739.
J. M. Harris, Richard Nolen‐Hoeksema, J. W. Rector, III, M. Van Schaack, and S. K. Lazaratos, (1992), "High resolution cross‐well imaging of a west Texas carbonate reservoir: Part 1. Data acquisition and project overview", SEG Technical Program Expanded Abstracts: 35-39.
Werner M. Heigl, Robert P. Radtke, Robert H. Stokes, and David A. Glowka, (2012), "Development of a downhole sparker source with adjustable frequencies", SEG Technical Program Expanded Abstracts: 1-5.
Larry Lines et al (1993), "Integrated reservoir characterization: Beyond tomography", SEG Technical Program Expanded Abstracts: 298-303.
Kimio Ogura, Takeichiro Ohhashi, Masaki Osada, and Masayoshi Yoshimura, (1992), "Design of a multidisk type downhole seismic source", SEG Technical Program Expanded Abstracts: 702-705.
Gildas Omnes, (1990), "Experimental study of the coupled cord downhole seismic source", SEG Technical Program Expanded Abstracts: 148-152.
Parker et al, (1994), “Presentation and discussion of tomographic inversions of crosswell surveys at Sellafield”, 56th EAEG Meeting, Jun 1994, cp-47-00262
Bjorn N. P. Paulsson, (1988), "Three‐component downhole seismic vibrator", SEG Technical Program Expanded Abstracts: 139-142.
Björn N. P. Paulsson, Robert P. Cutler, Glenn Kirkendall, Sen T. Chen, and John A. Giles, (1996), "An advanced seismic source for borehole seismology", SEG Technical Program Expanded Abstracts: 186-189.
Bjorn Paulsson, (2019), " A Fiber Optic Single Well Seismic System for Geothermal Reservoir Imaging & Monitoring”, 44th Workshop on Geothermal Reservoir Engineering Stanford University, Stanford, California, February 11-13, 2019
Nobusuke Shimada and Ymunori Shoji, (1997), "A crosswell measurement using by a multi‐stage, multi‐disk type downhole seismic source", SEG Technical Program Expanded Abstracts: 309-312.
G. A. Winbow, (1991), "Seismic sources in open and cased boreholes", GEOPHYSICS 56: 1040-1050.
Nalonnil, A., Marion, B., & Minto, J. (2013, March 26). Next Generation Borehole Seismic: Dual-Wavefield Vibrator System. International Petroleum Technology Conference. doi:10.2523/IPTC-16870-MS